1) if coinbase `unlock_time >= FCMP++ fork height`, then require
at least 1 coinbase out, and start doing FCMP++ torsion validation
2) Require FCMP++ txs' `reference_block+1 >= FCMP++ fork height`.
We want unique tree roots so that a tx with `reference_block=m`
can't possibly have a valid membership proof for some other
`reference_block=n`.
In RPC calls `get_blocks.bin` and `get_transactions`:
1. Skip deserializing then re-serializing tx blobs
2. Provide full unpruned transaction from mempool in `get_transactions` when `!req.prune`
3. Change `on_get_transactions()` response sort time from O(N^2) to O(N)
4. Generally improve memory usage and lookups in `on_get_transactions()`
Reviewed-by: selsta <selsta@sent.at>
Reviewed-by: j-berman <justinberman@protonmail.com>
Reviewed-by: jpk68 <jpk68@tutanota.com>
Simple API to check max blob size before parsing. This is useful
when reading blobs from untrusted sources.
@selsta pointed out that coinbase txs can technically be larger
than get_max_tx_size(), otherwise we could enforce it on all txs.
Never had a caller:
- Blockchain::init taking HardFork*&: added in 8f863e742, no caller in
any revision since
- tx_memory_pool::get_txpool_weight, set_txpool_max_weight: added as
get_txpool_size and set_txpool_max_size in bc61ae69b alongside
--max-txpool-size, which is wired through tx_memory_pool::init
instead; renamed size to weight in 5ffb2ff9b, still uncalled
- get_transaction_version: added in b750fb27b, no caller since
Dead when their last user went:
- core::get_blocks taking vector<pair<blobdata, block>>, both overloads:
last caller removed in 9faef1f83; ed2c81ed9 later converted the
already-dead signatures from std::list to std::vector. The
vector<block> overload is still used by core_tests and is kept
- blocks_container: 9e82b694d removed the in-memory blockchain that
used it
- tx_by_fee_and_receive_time_entry: 445319d3f replaced the sorted
container it keyed with a boost::bimap
- m_store_blockchain_interval: its do_call went with the original
blockchain_storage format in 9e82b694d
- m_fork_moaner: its do_call went with the "We are most likely forked"
message in f0371210e
Also drops two break statements after a return in ver_mixed_rct_semantics
and the doc comments belonging to the removed declarations.
- When there are missing IDs, blocks.size() != arg.blocks.size(), so arg.blocks can't be indexed by `i` - the indices will be wrong, the wrong weight will be returned to some peer and this peer will ban our node.
Use `bl.second` instead of `arg.blocks[i]`. Also it saves one DB query per returned block.
Does the following to speedup the `/is_key_image_spent` RPC endpoint:
- Reads all on-chain key images in one LMDB read transacion
- Uses `cryptonote_core::are_key_images_spent_in_pool()` instead of `cryptonote_core::get_pool_transactions_and_spent_keys_info()` for pool querying. This only does a LMDB read per key image if in the pool.
- Filters known on-chain spent key images before querying for key images spent in pool
This RPC endpoint was causing major daemon slowdowns in the Carrot/FCMP++ Alpha Stressnet when using the `rescan_spent` command, especially for large wallets.
The effect was much worse if the mempool was full.
